Anzeige
Erschienen in:
01.08.2014 | Article
Type 2 diabetes-related genetic risk scores associated with variations in fasting plasma glucose and development of impaired glucose homeostasis in the prospective DESIR study
Erschienen in: Diabetologia | Ausgabe 8/2014
Einloggen, um Zugang zu erhaltenAbstract
Aims/hypothesis
Genome-wide association studies have firmly established 65 independent European-derived loci associated with type 2 diabetes and 36 loci contributing to variations in fasting plasma glucose (FPG). Using individual data from the Data from an Epidemiological Study on the Insulin Resistance Syndrome (DESIR) prospective study, we evaluated the contribution of three genetic risk scores (GRS) to variations in metabolic traits, and to the incidence and prevalence of impaired fasting glycaemia (IFG) and type 2 diabetes.
Methods
Three GRS (GRS-1, 65 type 2 diabetes-associated single nucleotide polymorphisms [SNPs]; GRS-2, GRS-1 combined with 24 FPG-raising SNPs; and GRS-3, FPG-raising SNPs alone) were analysed in 4,075 DESIR study participants. GRS-mediated effects on longitudinal variations in quantitative traits were assessed in 3,927 nondiabetic individuals using multivariate linear mixed models, and on the incidence and prevalence of hyperglycaemia at 9 years using Cox and logistic regression models. The contribution of each GRS to risk prediction was evaluated using the C-statistic and net reclassification improvement (NRI) analysis.
Results
The two most inclusive GRS were significantly associated with increased FPG (β = 0.0011 mmol/l per year per risk allele, p
GRS-1
= 8.2 × 10−5 and p
GRS-2
= 6.0 × 10−6), increased incidence of IFG and type 2 diabetes (per allele: HR
GRS-1
1.03, p = 4.3 × 10−9 and HR
GRS-2
1.04, p = 1.0 × 10−16), and the 9 year prevalence (OR
GRS-1
1.13 [95% CI 1.10, 1.17], p = 1.9 × 10−14 for type 2 diabetes only; OR
GRS-2
1.07 [95% CI 1.05, 1.08], p = 7.8 × 10−25, for IFG and type 2 diabetes). No significant interaction was found between GRS-1 or GRS-2 and potential confounding factors. Each GRS yielded a modest, but significant, improvement in overall reclassification rates (NRI
GRS-1
17.3%, p = 6.6 × 10−7; NRI
GRS-2
17.6%, p = 4.2 × 10−7; NRI
GRS-3
13.1%, p = 1.7 × 10−4).
Conclusions/interpretation
Polygenic scores based on combined genetic information from type 2 diabetes risk and FPG variation contribute to discriminating middle-aged individuals at risk of developing type 2 diabetes in a general population.